Habitus from the Outside Unanswered Questions and the Presence of Institutions Nicolas Mariot1

Abstract This short paper aims to revisit what researchers do when they attempt, through questionnaires, interviews, or archives, to collect actors’ beliefs about highly regulated and institutionalized events (such as exams, ceremonies, official speeches, etc.). Three cases of non-response in surveys are examined in an attempt to interpret these silences. What do they indicate? At the end of this study, two points of view are advanced. The first is that collecting reasons is often a pointless exercise because it only serves to demonstrate the infinite diversity of actors’ reactions. The second is that it is problematic in the sense that the quest for beliefs leads the researcher to forget that, very often, the actors manage very well without them while still behaving “as they ought.” It is a question of activities where individual actions can be supported by social institutions.

1

Nicolas Mariot is a CNRS senior research associate (European Center for Sociology and Political Science, CESSP, Paris). His work focuses on the relationship between conformism and commitment in Western societies. His publications include, most notably, Bains de foule. Les voyages présidentiels en province, 1888–2002 (Paris: Belin, 2006), with Claire Zalc, Face à la persécution. 991 Juifs dans la guerre (Paris: Odile Jacob, 2010) and, most recently, Tous unis dans la tranchée ? 1914-1918, les intellectuels rencontrent le peuple (Paris: Seuil, 2013). http://www.jourdan.ens.fr/mariot/ ([email protected])

1

T

his short paper has come about as a result of an invitation from members of the

Politix editorial board to take another look at the controversies surrounding the role that is, and should be, assigned to beliefs in the analysis of political action. I would like to sincerely thank them specifically for this opportunity to revisit articles that were important during my own education and, more generally, because my discovery of the social sciences came just as the journal was publishing its first issues. As I told the organizers, I will say nothing here that I have not already said on the subject and, as you see, this written version preserves the “oral” style that it had originally. I simply want to take advantage of this anniversary to return to a specific problem related to the invitation given to me, and which I would formulate as follows: what is the use, in the social sciences, of studying beliefs? Beyond the response I will give, I will also use comparisons with other research to attempt to provide a somewhat broader scope than just my own areas of study. Why define the problem in this form? I want to use this question to look back at the discussions that took place in Politix around what I call the “normative model of the enlightened citizen.” In particular, I want to return to the back issues devoted to the socio-history of the vote and its frameworks2 and to those addressing the notion of political competency, especially “Le populaire et le politique I & II.”3 I think we could broadly summarize the contribution of these articles (they are far from being the only ones to have addressed the question but, since it is the anniversary, only this journal’s issues and articles will be mentioned) by saying that they came to gradually undermine the standard image of the citizen in democratic theory. This image was that of an individual as a social atom with a set of values, opinions, and representations that he was supposed to be able to express at all times, and which were meant to enable his behavior to be explained. This undermining remains incomplete in my view, largely due 2

J. L. Briquet and Y. Déloye, “La politique en campagnes,” Politix 15 (1991) and Y. Déloye and O. Ihl, “Des votes pas comme les autres,” Politix 22 (1993).

3

A. Collovald and F. Sawicki, “Le populaire et le politique,” Politix 13 and 14 (1991).

2

to the strength of the model. It is even possible to suggest that, more recently, among the many studies about (and for) “participative democracy,” it has sometimes taken a step backwards. The trend is quite real nonetheless. In some ways, my work simply consisted of lifting this analytical framework straight out of elections and applying it to the analysis of collective behaviors relating to official ceremonies. First point to note: you will see that I apply a very broad understanding of the notion of belief in this study. I do not mean it as something that is false or that people hold to be true against all the scientific evidence. I will not talk either about the difference between believing in (showing trust) and believing that (agreeing with a proposition).4 Here, I understand belief as referring broadly to ideas, thoughts, and opinions, to what people have in their heads, to that which is not visible but must be uncovered through questioning those being studied, in particular, using interviews and questionnaires or simply by having open discussions with them. I come back to my question then: what is the use, in social sciences, of studying beliefs? To attempt to respond to this, I would like to return to one of the first steps I took in my PhD program—a long time ago now, nearly as long ago as the first issues of Politix. During my first year in graduate school in 1992 to 1993, I chose François Mitterrand’s trip north for the inauguration of the new Euralille train station as my first study topic. The idea, in a very positivist approach, was to go and find out if what I had been told in different classes throughout my education—that political ceremonies reinforce social connection and revitalize the shared values of the group being considered—was true. This involved finding out whether the occasion could be measured using a questionnaire given out to the public. I designed a plan of action, intended to be as scientifically rigorous as possible. After lengthy negotiations, I obtained permission from the presidential palace to be included in the official cortège along with the necessary passes allowing me to follow the president. Five or six friends from my graduate course agreed to administer the questionnaire for me on the other side of the barriers. As agreed, they 4

J. Pouillon, “Remarques sur le verbe ‘croire’,” in La fonction symbolique, ed. M. Izard and P. Smith (Paris: Gallimard, 1979), 43–51.

3

were to ask every tenth spectator to fill out a simple form. On the front, there were questions, both closed and open of course, about the event and, on the back, there were questions collecting demographic information on the respondents. It was my intention that the analysis of the survey results would constitute the core of my master’s thesis. By asking the spectators about what they understood and how they felt about the ceremony’s staging, I remained straitjacketed by the dominant stimulus-response model. The goal for me was thus not to call this model into question, but simply to verify whether or not it could be empirically validated. This research logic seemed obvious to me. Since the function of political ceremonies (so I was told) is to revitalize the values of the assembled group, it seemed justified to go and ask the people whether or not they were in agreement with the actions and words of the speaker and even whether they could feel the force of the key symbols laid out in front of them, to use the usual terminology pertaining to this field.5 The underlying hypothesis, if I can still use this term, was that if a large enough number of participants acknowledged experiencing the same feelings or approving of the official statements being made, then I could conclude that the ceremony was effective and, if not, I could conclude that it was not. While this approach was comparable in some ways to asking worshipers at a church service if they believed in God or what message was conveyed in the sign of the cross they were making, I was not at all conscious of this when constructing the questionnaire. I did not yet understand that the specificity of this type of ceremony lies precisely in the fact that adherence to the gestures, symbols, and credo expressed is depersonalized because it is assumed collectively. The condition for its success lies in releasing those participants who wish to be released from any justification or explanation (they can take part while thinking of something else, in other words without believing in it). I was not yet familiar with the Langs’ article about MacArthur Day in Chicago,6 which was about one of the stops in the triumphal tour made by the general after his forced return from Korea. It is

5

See, for example, the classic work, M. Edelman, Politics as Symbolic Action (Chicago: Markham Publishing Company, 1971).

6

K. Lang and G. Engel Lang, “The Unique Perspective of Television and Its Effect: A Pilot Study,” American Sociological Review 18, no. 1 (1953).

4

still, even today, a model empirical study for the analysis of “mass approval ceremonies,” to use Robert Paxton’s formula.7 If I had been familiar with it, I would have realized the extent to which the behaviors and forms of the event are preestablished on both sides, by the organizers as well as the public or, to put it another way, the extent to which the social institution of the tour is essentially based on expectations, the absence of which would strongly disappoint the public. However, let us return to the Lille study. At first, the task appeared to me to be complete. Nearly three hundred questionnaires had been filled out and I had interesting demographic information about the public attending the event and at my disposal. Disappointment soon set in however. First of all, my friends who had administered the survey had told me about the respondents’ perplexity when they saw the questions. I noted that many of the questions had not been answered, in particular those where I had asked people to explain their attendance, to say what they thought of the event, or to give their opinion on specific elements of the ceremony. People had either basically repeated belief expressions when this had been present in the wording (a way of avoiding the issue, for example, yes, it is important to be there because it shows that we are united, because it is an important event, because after all it is the president, etc.) or, more often, they had given no response at all, often considering the questions to be incongruous. For a long time, I left this part of the survey aside, believing that I had worded things badly, rather than trying to understand what these silences said about the event and its function. For me, it was a technical problem to do with the survey, not the objectives. I would say now that the survey was not badly constructed (leaving aside the demographic section and the information it provided on the respondents); it was just meaningless. Let us now attempt to expand the scope of this very classic problem. Does studying an event that involves an audience mean having to understand how it is received by this audience? In other words, what are we doing when we give in, as I did, to the common temptation to say, does understanding the reality of a mass political event (a 7

R. Paxton, The Anatomy of Fascism (New York: Knopf, 2004).

5

demonstration, a meeting, the introduction of a new public policy measure) mean having to verify its effectiveness with the targeted audience? What I mean is, are we right to believe that the solution lies in going to find answers by asking people, for example, why are you here, why are you doing that (cheering, booing), what do you think about it, what is your opinion of . . . ? So far, I have mainly referred to surveys by questionnaire. However, the issue certainly does not just apply to the methods or techniques inherent in a survey. In this type of quest for respondents’ beliefs, opinions, values, or other internal representations, the fact that it is done using questionnaires, interview, or by examining archives makes no real difference to the core issue. In all cases, it is a matter of examining the personal motivations of those being studied, either directly, by asking them, (questionnaires or interviews) or indirectly, using documents generated by those who were in charge of monitoring or reporting on the event. This equivalence of methods is observable in the work done by historians on festivals and ceremonies and also, to a large extent, on demonstrations and other collective mobilizations. What does this work on the event’s reception by its audience involve for these researchers? Their end results amount to putting forward, using many boxes and stockrooms of archives, an inventory, which is as precise and complete as possible, of the variety of behaviors and attitudes in the audience.8 For Hazareesingh, the method (to carry out a thorough inspection of the departmental archives and festivities that are as representative as possible of territorial balances) was thus perfectly congruent with the goal of the study, which was to show that the citizen emerges under the influence of the empire. “This book is intended as a tribute to the political individualism of the French citizen,”9 he wrote in his introduction. Should we be surprised, however, that by increasing the number of 8

See, in particular, R. Dalisson, Les Trois couleurs, Marianne et l’Empereur. Fêtes libérales et politiques symboliques en France, 1815–1870 (Paris: La Boutique de l’Histoire, 2004) and S. Hazareesingh, The SaintNapoleon: Celebrations of Sovereignty in Nineteenth-Century France (Cambridge, MA: Harvard College). [S. Hazareesingh, La Saint-Napoléon. Quand le 14 juillet se fêtait le 15 août, trans. Guillaume Villeneuve (Paris: Tallandier, 2007)]

9

Hazareesingh, La Saint-Napoléon, 29. (Note: All citations from this work have been translated into English by the translator of this paper.)

6

observations, he ends up concluding that “Within each départment [French administrative area], there were localities where the Saint-Napoleon was warmly and enthusiastically celebrated, and others—sometimes just a few miles away—where the opposite was the case”?10 Above all, and we will come back to this point, is it really a decisive step forward to attest to the diversity that characterizes any collective practice? To qualify this kind of attempt (and once again, I think it is fairly generalized in sociological work on mobilizations), I will use the terms proposed recently by Boltanski contrasting “reality” with “the world.” The author starts out from the observation that, in sociology, “the possibility of radical uncertainty as to what things are really about” as well as “the anxiety it produces” are too often minimized or even unsatisfactorily masked. If we apply this to the field of festivals and ceremonies, we find that each one is subject to the risk of accident or failure. A consequence of the fragility of things, the description proposed by Boltanski of life in society leads him to reintegrate this radical uncertainty through the opposition between, on the one hand, the world, namely “everything that happens” that is uncertain and unpredictable and, on the other hand, reality, defined as that which “hangs together.”11 In this sense, uncertainty is connected to the world, while reality is the space where risk is measurable and controllable. We can say it yet another way: if the world represents that which is uncertain and hazardous in our existence, reality tends to provide stability and permanence. It is this “uncertain world” that, through critiques drawn from events, threatens the institutional arrangements of reality and makes it fragile.12 What is the point of this opposition for analyzing the types of political ceremony we have been discussing? We can show this by saying that, through looking first at individual appropriations of the ceremonial message, the specialist in the reception of events is not (or barely) interested in the material or mental reality of the rite as it is 10

Hazareesingh, La Saint-Napoleon, 16.

11

L. Boltanski, De la critique. Précis de sociologie de l’émancipation (Paris: Gallimard, 2009). (Note: All translations of citations from this work have been translated into English by the translator of this paper. No official English translation is in the public domain.)

12

L. Boltanski, De la critique, 91–95.

7

defined and pre-established by its institutions, particularly because these are viewed as limitations to the expression of the agents’ free creativity or deep feelings. What this specialist is trying to understand, from his inventory of personal reactions, is the world itself in all its diversity. However, this quest to encompass the vast world of participant reactions is, to my mind, subject to two potentially damaging risks. The first, which is almost unavoidable and already apparent in the case of SaintNapoleon, lies in being able to observe nothing more than an “infinite fragmentation of meanings.”13 We can obviously consider this risk to be negligible given the large number of studies whose principal result consists in describing, with endless lists as supporting evidence, the extraordinary variety of exhibited behaviors, depending on the different cases and circumstances. In my opinion, however, affirming this fragmentation does not necessarily represent an advance in knowledge. I think this risk is absolutely real and I think it is clear in the disappointment we feel when reading some studies of political rites. These may represent a lovely catalogue of festivals across the villages of France, but they leave us dissatisfied—this collection of beliefs that are like a stamp collection or a museum piece, like the treasures collected by anthropologists at the end of the nineteenth century. We also see it at work in historical studies of public opinion, such as those based on analyses of censored mail during the two world wars.14 The authors quote extracts from “pro” letters and then other sections from “anti” correspondence. Hidden in the middle of this juxtaposition is meant to be the unobtainable average opinion, but the inventory ultimately leads to the conclusion that there are “fors” and “againsts.”15 Lastly, we can see this danger in the recent popularity of attempts to pin 13

L. Boltanski, De la critique, 92.

14

See A. Cochet, “L’opinion et le moral des soldats en 1916 d’après les archives du contrôle postal” (PhD thesis, Université Paris X, 1986) and F. Lagrange, “Moral et opinions des combattants français durant la Première Guerre mondiale d’après les rapports du contrôle postal de la IVe armée” (PhD thesis, Université Paris IV, 2009). On Vichy, see the classic work by P. Laborie, L’opinion française sous Vichy. Les Français et la crise d’identité nationale, 1936–1944 (Paris: Seuil, 2001).

15

For more developed critical discussions, see A. Loez, “Pour en finir avec le ‘moral’ des combattants,” in Combats. Hommage à Jules Maurin, ed. J. F. Muracciole and F. Rousseau (Paris: Michel Houdiard, 2010) and B. Gaïti, “L’opinion publique dans l’histoire politique: impasses et bifurcations,” Le Mouvement Social 221 (2007).

8

down feelings and classify political emotions, the latest El Dorado in the sociology of mobilizations.16 As with studies on opinions, research faces a constant risk of exhaustion given the truly interminable (because it is infinite) nature of identifiable testimonies (without even discussing their solidity). The second risk inherent in the search for public reactions lies in remaining straitjacketed by the Western citizen norm. In the “social” perspective for analysis that I have been discussing, the goal is to reveal what remains hidden behind apparently shared decorum and behaviors. The event becomes worthy of interest when we can see in it the expression of multiple opinions, beyond the apparent homogeneity of attitudes. There is a shared objective for all these inventories, which, even in the most orchestrated of events (ceremony, elections), is that of individuals who retain free will, self-sufficiency, “autonomy of the will,” Eigensinn, and even “agency.” Second point to note: I want to point out in passing that this goal of reaffirming agents’ autonomy of the will can have very different political motivations. These range from illustrating the autonomy of the working classes even in festivals orchestrated and controlled by the dominant class—such as in Thompson’s analysis of charivari17—to showing that, even in the ex-Soviet-bloc countries (to cite but one example), ordinary people were less submissive than first appeared. However, we can see that, while the political continuum is broad, these efforts all tend in the same direction. They almost always attempt to reveal the autonomy behind the conformism and rarely show how resting on the social institutions is a much more frequent attitude than we think. This includes the most militant activists because they are the most accustomed to the machinery of collective action.

16

See the work by James Jaspers with titles evocative of this desire to rediscover agents’ creative autonomy: The Art of Moral Protest: Culture, Biography, and Creativity in Social Movements (Chicago: University of Chicago Press, 1997) and Getting Your Way (Chicago: University of Chicago Press, 2006) or his article “A Strategic Approach to Collective Action: Looking for Agency in Social Movement Choices,” Mobilization 9, no. 1 (2004).

17

E. P. Thompson, “‘Rough Music’: le charivari anglais,” Annales ESC 27, no. 2 (1972).

9

We can sum up the consequences of these two risks by exaggerating them slightly. Choosing to study the beliefs and opinions of participants means accumulating obstacles that block the way to perceiving how most of our actions are pre-established by social institutions, in particular when these are public and collective. It means not being able to see that we can participate without belief or even while thinking about or doing something else. This is particularly so because these are actions that go without saying, in other words they are meant to be accomplished without the need for participating individuals to justify them. I think that we end up here in a kind of internal contradiction inherent to an analysis of the “reception” of an institutionalized event. The researcher asks people to judge their own behavior, even though it is in fact preestablished and framed so as to free its agent from needing to formulate such a judgment or justification.18 When I was giving my oral presentation, many participants were looking at their cell phones. It is part of the nature of this type of public meeting to preserve a dimension of freedom by allowing those present to not fully participate. It is enough that the forms of the situation be respected overall, such as relative silence during presentations (but whispering is allowed because it could show interest or at least allow the speaker to think so), sometimes applause when a speaker is finished or at least a polite return to attention (listeners lift their heads and sit up in their seats), and a few questions or comments from those who are dedicated. If every research seminar or conference required absolute concentration on the part of its listeners for and about every topic, no one would run the risk of attending. From this perspective, I think that, what is decisive for social science, is not to inventory the immense possible variety of individual opinions and judgments, but rather to determine in what way the event is based on shared ideas and procedures that are never discussed and which allow it to be recognized. This is what I have tried to do by working on the process of depersonalization of instruments of celebration. The meaning of 18

For further development of this point, see F. Héran, “Le rite et la croyance,” Revue Française de Sociologie 27, no. 2 (1986).

10

gestures (applause, cheers, banners) preexists their expression; they in no way depend on what people have in their heads nor, in particular, on their intentions.19 Now I would like to expand the scope of the subject by leaving behind my ceremonial settings to look at other studies, which, to my mind, deal with the same sociological problem when confronted with cases of “non-response.” Among the possible choices, we could take a moment to look at the work of Orange. In two recent articles, the author takes a fresh look at the doctoral research in which she had tried to understand the reasons why students, mostly from the working and middle classes, chose advanced vocational training courses (post-baccalaureate courses leading to the senior technician diplomas [Brevet de technician supérieure – BTS]).20 A key aspect of the study involved passing out self-administered questionnaires in the classes. The students were asked to think about the choices they had made the previous year (their final year of high school) and also to explain the reasons behind their choice of further study. Orange soon saw that respondents regularly tripped up on a decisive question. It was the one asking them to recall the list of choices they had made during their final year of high school. Some respondents stumbled over this question and ended up abandoning it, saying “I don’t remember.” The author thus showed the existence of a very wide gap between the choices actually made during the final year of high school and those that the respondents were able to recall later. Nothing very surprising in that so far. Braconnier and Dormagen have shown how some of their study subjects could not remember their

19

For further development of this point, please refer to my earlier articles: “Qu’est-ce qu’un ‘enthousiasme civique’? Sur l’historiographie des fêtes politiques en France après 1789,” Annales. Histoire, Sciences Sociales 63, no.1 (2008) and “Does Acclamation Equal Agreement? Rethinking Collective Effervescence through the Case of the Presidential ‘Tour de France’ during the 20th Century,” Theory & Society 40, no. 2 (2001).

20

S. Orange, “Le choix du BTS. Entre construction et encadrement des aspirations des bacheliers d’origine populaire,” Actes de la Recherche en Sciences Sociales 183 (2010), and “Interroger le choix des études supérieures. Les leçons d’un ‘raté’ d’enquête,” Genèses 89 (forthcoming).

11

vote from one round of an election to the next.21 How then can we think that high school students might remember even more complex and distant choices? The situation got worse, however, when the author added, very logically, a question of opinion or judgment: “What were the two strongest motivations for your decision to pursue a BTS?” Several students dried up completely and were stuck. The researcher pushed them: “But still, you must have some idea . . . ” This got them no further and, after a long pause, they ended up answering: “Because you have to stay in school.” The author then showed that the problem raised by the non-responses is neither a question of method (there were items missing) nor entirely a lack of memory, but in fact a core sociological issue. This failure helps us see what is too often forgotten or what remains invisible when we “force” an answer. The act of choosing a course of study may not be a decision in the sense that it is often driven by social institutions (school, friends, family) and, furthermore, it is experienced as a requirement over which students have little influence (you have to stay in school). Contrary to what is assumed by requiring an individual and solitary response to a questionnaire, the question asked does not come from a decision made with full awareness and is not a solitary and personal act. In fact, Orange showed how the choice of study was, in many cases, not an individual choice because it may not have been a choice at all. The author described particularly illuminating scenarios where the decision was made among and with friends when it was time to fill out the forms in class, an exercise like any other, and that the head teacher consciously left the students alone with “their” decisions and responsibilities. I would like to finish by closing the circle, in fact by coming back to what is now an old issue of Politix, as old as my Lille questionnaire: the May 1993 issue entitled “Des votes pas comme les autres.” I would point out that, at that time of course, I was absolutely incapable of making the connection between the two. Why mention this issue? Because it contains two articles that raise, with evocative titles, questions very similar to those I have been asking and in a form that is doubtless more familiar to political scientists. 21

C. Braconnier and J. Y. Dormagen, La démocratie de l’abstention. Aux origines de la démobilisation électorale en milieu populaire (Paris: Gallimard, 2007).

12

These are the articles by Franceries, “Des votes aveugles. L’exemple des électeurs FN en milieu populaire” (Blind votes: The example of NF voters in the working classes), and by Daniel Gaxie, “Le vote désinvesti. Quelques éléments d’analyse des rapports au vote” (The disinterested vote: Some elements for analyzing relationships to voting). I cannot, in the space available, summarize the former, but simply wanted to note that it is based on an analysis of in-depth interviews with National Front voters and to quote the first two subsection headings, which are fairly explicit in relation to what I have just described and save me having to go into further detail: “Where we learn that voting is not a political gesture” and “Where it appears that voting requires no further comment.” In short, the author showed that, to understand why and how these voters chose to vote (because they did, in fact, vote) for the National Front, we cannot rely on what they said because they said nothing about it or perceived their action outside of the sphere of ideology. In relation to Gaxie’s article, I will again say just a few words directly related to my topic. The article is based on a survey by questionnaire at the exit of the polling stations during the 1989 municipal elections in Amiens. The author begins by pointing out that, of the 3,000 voters asked, fewer than half (1,353) agreed to respond. Of all the results obtained, I will settle on a single percentage which clearly calls to mind the questionnaire administered to the vocational students described above. More than a quarter (28.5%) of respondents gave no answer to the question of what counted the most for them in their voting decision. This percentage rose to 35% for those with no qualifications, manual workers, and those who were not very interested in politics, and even to 44% for those who declared themselves not at all interested in politics. This was true even though they had a significant selection of possible responses to choose from, which ranged from most to least political (services provided, personality of the leading candidate, composition of the list, a general issue, the state of the city, accomplishments, program, political considerations, other responses). I simply want to comment on these few results by adding that, while we are right to focus on the investment differential based on class, this should not lead us to forget the significance of the average rate (more than a quarter of non-responses). In every social category, the 13

phenomenon of resting on social institutions exists. Moreover, giving a response does not necessarily mean a strong and ongoing commitment. Recent studies have shown, in addition, that disinterested or collective voting can be a socially shared activity, such as among Web users in the 16th arrondissement of Paris (taking, for example, the Paris en campagne survey).22 By way of conclusion, we can suggest that, on the basis of these three examples of partially incomplete questionnaires, we simply have to bear in mind that the nonresponses are not just a technical problem to do with the formulation of the questions,23 nor are they simply a problem restricted to responses from the working classes. The regularity with which social science surveys on a whole range of subjects register “blank” responses such as these raises a much broader sociological question that I think too often goes unnoticed. It is the fact that, in many areas of our societal lives, our actions seriously lack commitment. To the question, “what is the use, in social sciences, of studying beliefs?” I would therefore respond (jokingly) that the process should at least show that, for many social activities, people do not have any beliefs and that they manage very well without them. Beyond such evasions, this conclusion seems to me to be important for thinking more deeply than we usually do about the role played by conformism in our societies. One possible way to develop studies of this topic could thus consist in systematically examining the material systems and “institutions of meaning” attached to each of our social activities and the fact that we mobilize them without paying them any attention. Both share an interesting quality in that they are immediately observable. It is in this sense that these pre-established ways of doing and thinking make up what I called in the title a “habitus from the outside”: their power is external to individuals and therefore does not depend on what people think of them.

22

See É. Agrikoliansky and S. Lévêque, “Les absents du scrutin: logiques de la démobilisation électorale” and É. Agrikoliansky, J. Heurtaux, and B. Le Grignou, “Des conduites sans croyance? Mobiliser dans les ‘beaux quartiers’” in Paris en campagne. Les élections municipales de mars 2008 dans deux arrondissements parisiens, ed. É. Agrikoliansky, J. Heurtaux, and B. Le Grignou (Bellecombe-en-Bauges: Éditions du Croquant, 2011).

23

On this point, see the work of J. P. Grémy, “Questions et réponses: quelques résultats sur les effets de la formulation des questions dans les sondages,” Sociétés Contemporaines 16 (1993).

14